- Email: [email protected]
Neuromyelitis optica testing and treatment: Availability and affordability in 60 countries
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
Contents lists available at ScienceDirect
Multiple Sclerosis and Related Disorders journal homepage: www.elsevier.com/locate/msard
Neuromyelitis optica testing and treatment: Availability and affordability in 60 countries
T
Kathryn Holroyda, Andre Vogela, Kat Lynchb, Brittany Gazdagb, Matthew Voghelb, ⁎ Nicholas Alakelc, Bryan N. Patenauded, Horacio Chiong-Riveroa, Farrah J. Mateena, a
Department of Neurology, Massachusetts General Hospital, 55 Fruit St., Boston, MA 02114, USA ZoomRx, 521 Broadway, New York, NY 10012, USA c ZoomRx, 50 Milk St., Boston, MA 02109, USA d Bloomberg School of Public Health, Johns Hopkins University, 615N. Wolfe Street, Baltimore, MD 21205, USA b
ARTICLE INFO
ABSTRACT
Keywords: Neuromyelitis optica Epidemiology Global health Health economics Immunosuppressive therapy
Background: We characterize the variations in availability and affordability of NMO diagnostic testing and treatment by geographic region and country-level income group. Methods: A structured survey was distributed in English, French, and Spanish in late 2018 to neurologists and other physicians who encounter NMO patients. Results: Respondents (response rate 45%, 64/143 countries contacted) came from all WHO world regions and World Bank country income levels (49% university-based; 13 low-, 16 lower middle-, 16 upper-middle-, and 15 high-income countries). The average cost of an aquaporin-4 antibody (AQP4-Ab) test to a patient globally was 209 USD, and the average cost of NMO treatment per year was 3,819 USD. AQP4-Ab and myelin oligodendrocyte glycoprotein-antibody (MOG-Ab) testing were available in 68% and 38% of all countries. Low-income countries had poor availability of both AQP4-Ab (2/13 countries) and MOG-Ab (1/13) compared to high-income countries (15/15 AQP4-Ab, 13/15 MOG-Ab). Nearly half (48%, 13/27) of African and Eastern Mediterranean countries had access to neither test. Global treatment availability and usage: Azathioprine (88%), rituximab (50%), mycophenolate mofetil (57%), intravenous methylprednisolone (98%), oral prednisone (68%), plasma exchange (78%), intravenous immunoglobulin (72%). Whereas 70–100% of high-income countries’ patients could afford treatment without incurring a catastrophic health expenditure, <10% of low-income country patients could. Most low-income countries (12/13) reported the patient pays for NMO care entirely without public assistance Conclusions: There is a gap in access to diagnostic testing for NMO in non-high-income countries, even in countries where acute and immunosuppressive treatment for NMO are available.
1. Introduction Neuromyelitis Optica (NMO) is found in diverse populations with a prevalence ranging from 0.5 to 4.4 cases per 100,000 people. (CabreraGómez et al., 2009; Asgari et al., 2011; Etemadifar et al., 2015) However, the burden of NMO is likely higher than recognized, especially in resource-limited settings, given variations in knowledge and access to diagnostic tests as well as the misdiagnosis of NMO as multiple sclerosis (MS). (Mealy et al., 2018) NMO differs in treatment from MS, requiring diagnostic distinction at the time of presentation. In 2017, the International Panel for NMO Diagnosis convened to create uniform
diagnostic criteria. (Wingerchuk et al., 2015a) In addition to clinical characteristics, the criteria rely heavily on neurodiagnostic tests, including serum antibody testing and MRI findings. Diagnostic tests for NMO are anecdotally known to be unavailable in several locations globally. In a 2016 study, the aquaporin-4 antibody (AQP4-Ab) test was reported as the most commonly sought after but unavailable test in SubSaharan African postgraduate neurology training programs. (Mateen et al., 2016) NMO is also not a benign condition, with a disease mortality ranging from 9% to 32% in a high-income country, (Mealy et al., 2018) making diagnostic barriers potentially fatal. Early NMO diagnosis and immunosuppressive treatment are
Correspondence to: Department of Neurology, Massachusetts General Hospital, 165 Cambridge Street, #627, Boston, MA 02114, USA. E-mail addresses: [email protected] (K. Holroyd), [email protected] (A. Vogel), [email protected] (K. Lynch), [email protected] (B. Gazdag), [email protected] (M. Voghel), [email protected] (N. Alakel), [email protected] (B.N. Patenaude), [email protected] (H. Chiong-Rivero), [email protected] (F.J. Mateen). ⁎
https://doi.org/10.1016/j.msard.2019.05.013 Received 21 March 2019; Received in revised form 8 April 2019; Accepted 20 May 2019 2211-0348/ © 2019 Elsevier B.V. All rights reserved.
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
essential to reduce both morbidity and mortality for patients. (Pandit et al., 2013; Trebst et al., 2014) In higher income settings, case fatality is thought to have decreased since the introduction of the AQP4-Ab test in 2004. (Collazo et al., 2018) It is unclear if this benefit has reached most populations. We are unaware of global studies that have characterized physicians’ familiarity or the availability and affordability of NMO diagnosis and treatment, although others have examined the characteristics of NMO in lower-income country populations. (Gaye, 2018; Assadeck et al., 2019; Bhigjee et al., 2017) We sought to understand the patterns of diagnosis and treatment of NMO worldwide by surveying a convenience sample of physicians globally who encounter NMO patients.
The survey and introduction were written and distributed in English, French, or Spanish, with the option to request a translation into other languages as the first question of the survey (Appendix A, survey instruments). A gift card for 50USD was offered upon survey completion. 2.5. Data analysis When multiple respondents from a country completed the survey, the first complete submitted response was used in the analysis. Countries were stratified by WHO geographical region (https://www. who.int/about/regions/en/) and World Bank income status from the most recently available year World Bank Data Team (2018). Currencies were converted to USD using the exchange rates on xe.com on November 8, 2018. Affordability analyses were conducted using Stata (v.15.1, College Station, TX).
2. Methods 2.1. Ethics approvals
2.6. Cost analysis
The Partners Healthcare Institutional Review Board exempted this study.
To generate estimates of the affordability of NMO immunosuppressive medications in each country, the distribution of household income was extrapolated. Given limited comparable data on household income by percentile, income distributions were constructed using per capita gross domestic product (measured in purchasing power parity adjusted constant 2011 international$) and the most recent Gini index values, a measure of income inequality, obtained from the World Bank Databank. The precise method for constructing the estimated income distributions has been reported (Harttgen and Vollmer, 2011). In order to adjust the estimated distribution to reflect a household's disposable income, rather than purely individual aggregate income: (1) labor's share of income was assumed to be 2/3 of total income, an estimate stemming from economic theory (Kaldor, 1961; Solow, 1957; Johnson, 1954); (2) it was assumed that there were two wage earners per household; (3) household subsistence was defined as the average number of members per household in each country multiplied by the international annual poverty line. The average number of members per household for each country was obtained from the UN Statistical Database. The international subsistence level was defined as an individual living on $693.50 PPP-adjusted 2011 constant $ per year ($1.90 per day) by World Bank standards (Ravallion et al., 2008). This $693.50 per year is multiplied by the average number of household residents to calculate a measure for household subsistence in constant dollars. We assume that until household income surpasses this bare minimum, there is no disposable income. Household subsistence was subtracted from the previously generated income distribution to calculate household disposable incomes. Affordability was assessed by calculating the cost of one year of NMO treatment as a percentage of annual household disposable income. The WHO defines medical expenditures that are >40% of household disposable income as catastrophic. However, even an expenditure <40% of annual household income is likely catastrophic for most families in the world.
2.2. Survey design and implementation A new survey instrument was designed in 2018 by the authors who include neuroimmunologists, neurology residents, and researchers proficient in survey design for physicians. The survey was piloted on a small group of neuroimmunology-focused neurologists to improve the clarity and sequencing of the questions. The final survey instrument was edited for format, content, and design and had up to 48 multiple choice and fill-in-the-blank questions with contingencies based on prior responses. The goal of the survey was to understand: 1) practice and institution information of the respondent; 2) familiarity with the medical aspects of NMO; 3) imaging available for diagnosing NMO and barriers to its acquisition; 4) diagnostic laboratory tests available for diagnosing NMO and barriers to obtaining these tests; and 5) treatments available for NMO in the acute and chronic phases. The survey was distributed on an independent survey platform, hosted by the company ZoomRx, and available to potential respondents from July to November 2018. 2.3. Respondent selection Countries with >1 million inhabitants and designated as low- or middle-income by the World Bank were prioritized. (Bhigjee et al., 2017) Medical providers thought to be familiar with NMO were identified through several routes: (1) PubMed was searched with MeSH terms, including the country name of interest AND [“neuromyelitis optica,” OR “multiple sclerosis,” OR “neuroimmunology”], and if no findings, the country name AND “neurology;” (2) An online search of neuroimmunologists on faculty at academic hospitals in countries of interest was performed; (3) Delegates of international neuroimmunology committees were located on publicly accessible websites; (4) Presenting authors from countries of interest were found through international neuroimmunology conferences; and (5) Professional contacts of the physician authors’ networks were contacted. In countries where there were no subspecialists of neuroimmunology found, general neurologists were identified. In countries where no neurologists were identified, non-neurologist physicians who would take care of NMO patients were identified.
2.7. Data availability Anonymized survey data will be made available to qualified investigators upon request. 3. Results
2.4. Survey implementation
3.1. Respondents
The response goal was one respondent per country. When a country had a respondent, no further respondents were sought from that country. Respondents were emailed with an introductory message of standardized text requesting their participation in the survey, with a reminder email sent 1 to 4 weeks later if there was no initial response.
Of the 193 UN member states, a physician in 143 was contacted. Responses were received from 60/143 (42%) of countries contacted. Additional respondents from four other countries reported unfamiliarity with NMO and were removed from further analyses (overall response rate 64/143 countries=45%). 45
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
institution. 71% of high-income country respondents practiced at an academic institution, while 39% of low-income and 40% of lowermiddle-income country respondents did so. By WHO region, 67% of European respondents reported practicing at an academic institution compared to only 39% of respondents from African countries.
Table 1 Respondent characteristics. n = 60 Degree of experience with NMO diagnosis and treatment: Familiar, but never diagnosed Familiar, have diagnosed, but never treated Fairly familiar, have diagnosed, treated, but not often Very familiar, diagnose and treat regularly Setting of primary practice location: Rural (<50,000 residents) Smaller city (100,000 to 1 million residents) Larger city (100,000 to 1 million residents) Urban area (>1 million residents) World Bank income status of country of respondent: Low income (L) Lower-middle income (LM) Upper- middle income (UM) High income (H) WHO region of respondent: Africa (Af) Americas (Am) Eastern Mediterranean Region (EM) Europe (Eu) South-East Asia (SEA) Western Pacific (WP) Primary treated age group: Children (0–17 years) Adults (18+ years) Both children and adults Confidence treating NMO. 1 = Not at all, 7 = extremely confident 1–3 4 5–7
7 (12) 4 (7) 23 (38) 26 (43)
3.2. Diagnostic tests Across all country-level respondents, AQP4-Ab and MOG-Ab were present in 41 (68%) and 23 (38%) countries respectively. 15 of 15 (100%) high-income countries versus 2 of 13 (15%) low-income countries reported AQP4-Ab availability. 13 of 15 (87%) high-income countries versus 1 of 13 (8%) low-income countries reported MOG-Ab availability. 13 of 27 (48%) respondents in Africa and the Eastern Mediterranean regions reported that the AQP4-antibody test is unavailable. In contrast, only 6 of 33 (18%) respondents from countries of any income level in the pooled Americas, Europe, South-East Asia, and the Western Pacific reported that the AQP4-Ab test is unavailable. Of the 41 respondents who reported availability of the AQP4-Ab test, 20 (49%) stated they had access to the test without issue. Similarly, of the 23 respondents who reported availability of the MOG-Ab test, 13 (57%) reported having access to the test without issue. However, when respondents were aggregated by WHO world region, only 4 of 14 (29%) respondents from Africa and the Eastern Mediterranean reported access to the AQP4-Ab test and 2 of 8 (25%) reported access to MOG-Ab test without issue (Table 2). Most respondents had access to the other diagnostic tests that support the diagnosis of NMO. Fifty respondents (83%) reported availability of a CT scanner on site, 42 (68%) had an MRI scanner on site, 50 (83%) could perform a lumbar puncture on site, 36 (60%) could access visually evoked potentials on site, and 34 (57%) could access optical coherence tomography on site. Only CT scans and lumbar punctures were available on site in most low-income countries (Fig. 2). Countries in Africa and the Eastern Mediterranean regions reported that MRI, CT, and lumbar puncture were available at roughly similar rates as countries in the Americas, Europe, South-East Asia, and the Western Pacific; however, they reported significantly lower rates of availability of visual
1 (2) 2 (3) 14 (23) 43 (72) 13 16 16 15
(22) (27) (27) (25)
15 (25) 10 (17) 12 (20) 10 (17) 6 (10) 7 (12) 2 (3) 50 (83) 8 (13) 4 (7) 10 (17) 46 (77)
Of the 60 respondents who completed the survey, >90% of respondents had at least an average level of confidence in their care of people with NMO (Table 1). Respondents from all world regions and country-income level strata completed the survey (Fig. 1). Most respondents practiced in an urban area of >1 million inhabitants with only one respondent practicing in a rural area of <50,000 inhabitants. Approximately half (49%) of respondents practiced at an academic
Fig. 1. Map of responding countries by World Bank income status. 46
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
Table 2 Availability, access, and barriers to NMO diagnosis and treatment across income levels and world regions. Availability and access of diagnostic tests
H (n = 15)
UM (n = 16)
LM (n = 16)
Availability of AQP4-Ab diagnostic 15 (100) 13 (81) 11 (69) test Availability of MOG-Ab test 13 (87) 4 (25) 5 (31) No availability of either AQP4- or 0 (0) 3 (19) 5 (31) MOG-Ab tests Where AQP4-Ab test available, 11/15 (73) 4/13 (29) 3/11 (23) access without issue (n = 44) Where MOG-Ab test available, 10/13 (77) 1/4 (25) 1/5 (17) access without issue (n = 24) On-site and referral availability of common NMO diagnostic tools CT scan 13 (87) 16 (100) 15 (94) Lumbar puncture 14 (93) 16 (100) 13 (81) Optical Coherence Tomography 14 (93) 13 (81) 7 (56) MRI 15 (100) 16 (100) 14 (88) Visually Evoked Potentials 14 (93) 12 (75) 11 (69) Availability and use of long-term NMO treatments Azathioprine 13 (87) 15 (94) 16 (100) Mycophenolate mofetil 13 (87) 9 (56) 9 (56) Rituximab 12 (80) 9 (56) 6 (38) Challenges to treatment of NMO Inability to consistently obtain 1 (7) 8 (50) 8 (50) medicine supplies Cost of medication to health sector 6 (40) 11 (69) 8 (50) Cost of medication to patient 1 (7) 11 (69) 12 (75) Practice lacks access to preferred 4 (27) 6 (38) 6 (38) medications
L (n = 13)
Af (n = 15)
Am (n = 10)
EM (n = 12)
Eu (n = 10)
SEA (n = 6)
WP (n = 7)
2 (15)
7 (47)
7 (70)
7 (58)
9 (90)
5 (83)
6 (86)
1 (8) 11 (85)
3 (20) 8 (53)
3 (30) 3 (30)
5 (42) 5 (42)
7 (70) 1 (10)
1 (17) 1 (17)
4 (57) 1 (14)
2/2 (100)
2/7 (29)
3/7 (43)
2/7 (29)
7/9 (78)
2/5 (40)
4/6 (67)
1/1 (100)
1/3 (33)
1/3 (33)
1/5 (20)
6/7 (86)
1/1 (100)
3/4 (75)
8 9 5 7 5
(62) (69) (38) (54) (38)
11 (73) 12 (80) 6 (40) 10 (67) 5 (33)
10 (100) 10 (100) 9 (90) 10 (100) 8 (80)
10 (83) 10 (83) 6 (50) 10 (83) 10 (83)
9 (90) 10 (100) 9 (90) 10 (100) 9 (90)
6 5 5 5 4
6 5 6 7 6
9 (69) 3 (23) 3 (23)
11 (73) 7 (47) 2 (13)
9 (90) 6 (60) 6 (60)
12 (100) 5 (42) 7 (58)
9 (90) 9 (90) 8 (80)
6 (100) 3 (50) 3 (50)
6 (86) 4 (57) 4 (57)
9 (69)
11 (73)
4 (40)
5 (42)
1 (10)
3 (50)
2 (29)
5 (38) 11 (85) 9 (69)
7 (47) 13 (87) 11 (73)
7 (70) 5 (50) 4 (40)
8 (67) 8 (67) 7 (58)
4 (40) 1 (10) 1 (10)
2 (33) 5 (83) 1 (17)
2 (29) 3 (43) 1 (14)
(100) (83) (83) (83) (67)
(86) (71) (86) (100) (86)
Note: H = High-income countries, UM = Upper middle-income countries, LM = Lower middle-income countries, L = Low-income countries. Af = Africa, Am = Americas, EM = Eastern-Mediterranean, Eu = Europe, SEA = South-East Asia, WP = Western Pacific.
Fig. 2. Availability and accessibility of non-antibody diagnostic tests by A) World Bank income status and B) WHO World Region of responding countries. 47
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
Fig. 3. Availability and accessibility of long-term NMO treatments by A) World Bank income status and B) WHO World Region of responding countries.
evoked potentials and optical coherence tomography. Respondents from 28 countries reported an estimate of the price of AQP4-Ab testing. The average reported price for the AQP4-Ab test was 209.29USD (median 91 USD).
3.4. Perceived barriers to NMO care The most commonly reported challenge was the cost of immunosuppressive medication to the patient, followed by an inability to obtain a supply of medications, and the cost of the medication to the health sector. A minority (7, 12%) of respondents reported never having experienced challenges treating NMO. With the exception of respondents from high-income countries, the most commonly reported challenge in treating NMO was the cost of medicines to the patient. A majority (11, 85%) of low-income countries found patient-borne costs were a challenge in treatment, while only 1 of 15 (7%) high-income countries reported the same. For respondents from high-income countries, the greatest challenges in treating NMO were the cost to the health sector, inadequate patient understanding of the disease, and patient non-adherence. 9 (69%) respondents from lowincome countries reported lack of access to medications as a barrier to treatment. In contrast, only 4 (27%) respondents from high-income countries reported that care is impeded by a lack of access to medications.
3.3. Treatment Access to treatment for acute NMO relapses was high: 59 (98%) respondents reported the availability of intravenous methylprednisolone, 41 (68%) of oral steroids, 47 (78%) of plasma exchange, and 43 (72%) of intravenous immunoglobulin. Availability of acute NMO treatments increased with a country's World Bank income status from low to high. Access to long term NMO immunosuppressive treatments was variable. Most (n = 53 of 60, 88%) respondents stated that azathioprine is both available and used. Use of mycophenolate mofetil and rituximab increased with increasing country-level income group. Mycophenolate mofetil and rituximab were available in 3 low-income countries (23%) compared to high-income countries where 13 (87%) and 12 (80%) had availability of mycophenolate mofetil and rituximab respectively (Fig. 3). For patients in whom it is impossible to distinguish between NMO and MS at initial presentation based on available in-country resources, there was variation in treatment approach. Clinicians would treat as NMO (25, 42%), MS (16, 27%), or not treat for either disease (19, 32%). Of those who would treat a patient as MS when uncertain on the diagnosis, half of the respondents (8, 50%) stated they would consider azathioprine for initial treatment. Other frequent medication choices included interferon beta-1a, glatiramer acetate, natalizumab, and rituximab.
3.5. Affordability The average annual cost to a typical patient for NMO treatment was estimated at 3818.56 USD (n = 52 responding countries). 21 (35%) respondents stated that a patient's costs are not typically covered by public health insurance, 18 (30%) stated that a patient's costs are partially covered, and 20 (33%) stated that a patient's costs are covered in full (n = 1 no answer). All 15 high-income countries reported that patients’ treatment costs were covered partially or fully by public insurance. In contrast, 12 (92%) low-income countries reported that a patient's costs are not covered at all by public insurance, with one low48
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
Fig. 4. Predicted populations percentages able to afford 1 year of NMO treatment by expending <40% or <10% of annual household income.
consensus criteria for NMO rely heavily on the availability of MRI and blood tests which may be difficult to obtain and require the mailing out of blood samples to laboratories out of country with prolonged waiting time for results. For all countries to reach the same standard of care for the diagnosis of NMO, more universal access to diagnostic tests is necessary. With a more affordable, accessible point of care diagnostic test for NMO, for instance, the ability to diagnose and therefore treat NMO would likely be greatly increased internationally. Since the NMO diagnostic criteria are understandably dependent on the sensitive and specific biomarkers that are available, it is debatable whether the criteria should be modified for level of certainty in low-resource settings versus the diagnostic tests themselves should be made more available, accessible, and affordable. Presently, many patients’ out of pocket costs for diagnosis can far exceed the cost of treatment with azathioprine. For health systems that do not have access to or cannot afford diagnostic testing, the question of treating the disease as NMO or as multiple sclerosis remains unsettled. Although we hypothesized that access to both diagnosis and treatment are limited in more resource-limited settings, our respondents most often reported limitations in NMO diagnosis. To underscore this dilemma, we found that 69% of low-income country respondents had access to azathioprine while only 15% reported access to AQP4-Ab and 8% access to MOG-Ab testing. The ability to treat patients using prednisone and azathioprine was relatively higher than expected. Since azathioprine is available on the WHO's Essential Medicines Lists for other indications, the use of azathioprine is usually possible. Meanwhile, the treatment landscape for NMO continues to expand in high-income countries. For instance, one novel therapy, eculizumab, is currently estimated to cost ∼500,000 USD per patient per year (Pittock et al., 2013; Dinerstein, 2017). For this same price, we estimated that >2000 AQP4-Ab tests could be performed, using the average reported cost in our study of 209.29 USD. There are several limitations to our study. The response rate is 45%. Although this is low compared to the number of countries we hoped to reach, it is similar to other published surveys reaching lower income countries (McLane et al., 2015) and includes a fairly even representation across geographic and country income levels. Countries in lower
income country reporting that a patient's costs are partially covered. In low-income countries for which disposable household income assessment was possible due to available demographic data, none of the patient population could afford the cost of 1 year of NMO medication by spending <10% of disposable income. Less than 10% of the population in any low-income country (often 0), could afford 1 year of NMO medication without incurring catastrophic (>40% of household disposable income) health expenditures. By contrast, 70 to 100% of the population in high-income countries analyzed could afford one year of NMO immunosuppression without incurring catastrophic health expenditures (Fig. 4). 4. Discussion We present data from a global convenience sample that synthesizes practitioners’ ability to diagnose and treat NMO across geographic and economic zones. Whereas prior reports of a global nature have considered the landscape of neurological care by the type of diagnostic tests available (McLane et al., 2015) or the global burden of a particular neurological disease, (Zunt et al., 2016; Stovner et al., 2018; Feigin et al., 2017) few studies have combined these two approaches. Our study allows an improved understanding of the delivery of care for a neuroimmunological disease and how the availability of testing and treatment varies across broad world regions. Since NMO has relatively recently been defined by its distinct biomarkers (Wingerchuk et al., 2015b) our study also sheds light on the pace of uptake of new technologies and treatments for a neurological disease around the world when they originate from a high-income setting. Our study both emphasizes existing trends in health systems and sheds new light on the burden neurologists face when encountering a less common neurological disorder. Our results, although mainly descriptive, could lead to several important interventions that improve the lives of patients with NMO and enable the practitioners who encounter them. First, we identified several countries where neurology practitioners, whether neurologists or not, were unfamiliar with NMO and could be invited to ongoing educational initiatives to increase NMO awareness. Second, diagnostic 49
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
income settings may have been more motivated to reply, because they have recognizable needs and fewer available respondents. It is possible that we overlooked eligible clinicians in some countries. Because we reached out to personal contacts, our respondents’ demographic characteristics may be biased towards those in a similar practice type as the authors, i.e. in academic settings. Our results are subject to selection bias. Since our a priori response goal was to achieve one key informant in a country, we did not aim to have multiple respondents which could have provided a more complex picture of NMO care within each country. Although survey translation was offered, the survey was initially distributed in English, with French and Spanish translations made available following requests. This limited our response rate to people who spoke at least basic levels of one of these languages. The use of a web-based survey may have led to selection bias towards respondents with better access to other resources; however, we could not find any other way to reach so many practitioners globally in such a short period of time. Responses may have been inaccurate and questions may have been misunderstood. We report trends and patterns, but do not identify causal associations. It is well-recognized that countries have significant regional variability and sub-national inequities. Our respondents, being disproportionately in urban, academic centers, may provide better than usual care in a country. Some countries have insufficient baseline household information to calculate affordability because they are too poor or fragile. Given this, the affordability of NMO diagnosis and treatment may have been over-estimated. Much work remains to be done to increase the availability and affordability of diagnostic tests for neuroimmunological diseases like NMO in lower income countries. Although treatments for acute disease relapses are generally available, very limited access to testing can be used to direct management. Increasing international education programs and collaborating with established non-governmental organizations and pharmaceutical companies could help increase awareness for NMO and access to existing diagnostic tests, or even contribute to development of a point-of-care test.
References Asgari, N., Lillevang, S.T., Skejoe, H.P., Falah, M., Stenager, E., Kyvik, K.O., 2011. A population-based study of neuromyelitis optica in Caucasians. Neurology 76 (18), 1589–1595. Assadeck, H., Toudou Daouda, M., Adehossi Omar, E., Mamadou, Z., Hassane Djibo, F., Douma Maiga, D., 2019. Inflammatory demyelinatiung diseases of the central nervous system in Niger. Rev. Neurol. (Paris). 175 (4), 261–268. Bhigjee, A.I., Moodley, A.A., Roos, I., Wells, C.L., Ramdial, P., Esser, M., 2017. The neuromyelitis optica presentation and the aquaporin-4 antibody in HIV-seropositive and seronegative patients in KwaZulu-Natal, South Africa. South Afr. J. HIV Med. 18 (1), 684. Cabrera-Gómez, J.A., Kurtzke, J.F., González-Quevedo, A., Lara-Rodríguez, R., 2009. An epidemiological study of neuromyelitis optica in Cuba. J. Neurol. 256 (1), 35–44. Collazo, I.M., Weinshenker, B., Nasr, Z., Lopez, A., 2018. Long term mortality in neuromyelitis optica spectrum disorder (P6.414). Neurology 90 (April(15 Supplement)). Dinerstein, C., 2017. Why is Soliris the Most Expensive Drug in the US? American Council on Science and Health [Internet][Accessed March 7, 2019]. Available from: https:// www.acsh.org/news/2017/05/27/why-soliris-most-expensive-drug-us-11333. Etemadifar, M., Nasr, Z., Khalili, B., Taherioun, M., Vosoughi, R., 2015. Epidemiology of neuromyelitis optica in the world: a systematic review and meta-analysis. Mult. Scler. Int. 2015, 174720. Feigin, V.L., Abajobir, A.A., Abate, K.H., Abd-Allah, F., Abdulle, A.M., Abera, S.F., et al., 2017. Global, regional, and national burden of neurological disorders during 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet Neurol. 16 (November (11)), 877–897. Gaye, N.M., Fall, M., Diop, A.M., 2018. Neuromyelitis optica spectrum disorders (NMOSD) in a Sub-Saharan Africa country: a preliminary study of sixteen Senegalese cases. Mult. Scler. Relat. Disord. 27, 179–183. Harttgen, K., Vollmer, S., 2011 November. United Nations Development Programme. Human Development Reports. Research Paper 2011/13. Johnson, D.G., 1954. The functional distribution of income in the United States, 18501952. Rev. Econ. Stat. 36 (2), 175–182. Kaldor, N., 1961. Capital accumulation and economic growth. The Theory of Capital. Macmillan & Co Ltd. Mateen, F.J., Clark, S.J., Borzello, M., Kabore, J., Seidi, O., 2016. Neurology training in sub-Saharan Africa: a survey of people in training from 19 countries. Ann. Neurol. 79 (6), 871–881. McLane, H.C., Berkowitz, A.L., Patenaude, B.N., et al., 2015. Availability, accessibility, and affordability of neurodiagnostic tests in 37 countries. Neurology 85 (18), 1614–1622. Mealy, M.A., Kessler, R.A., Rimler, Z., et al., 2018. Mortality in neuromyelitis optica is strongly associated with African ancestry. Neurol. Neuroimmunol. Neuroinflamm. 5 (4), e468. Pandit, L., Mustafa, S., Kunder, R., Shetty, R., Misri, Z., Pai, S., 2013. Optimizing the management of neuromyelitis optica and spectrum disorders in resource poor settings: experience from the Mangalore demyelinating disease registry. Ann. Indian Acad. Neurol. 16 (4), 572–576. Pittock, S.J., Lennon, V.A., McKeon, A., et al., 2013. Eculizumab in AQP4-IgG-positive relapsing neuromyelitis optica spectrum disorders: an open-label pilot study. Lancet Neurol. 12 (6), 554–562. Ravallion, M., Chen, S., Sangraula, P., 2008. Dollar a Day Revisited (Report). The World Bank, Washington DC. Solow, R.M., 1957. Technical change and the aggregate production function. Rev. Econ. Stat. 39 (August (3)), 312. Stovner, L.J., Nichols, E., Steiner, T.J., Abd-Allah, F., Abdelalim, A., Al-Raddadi, R.M., et al., 2018. Global, regional, and national burden of migraine and tension-type headache, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 17 (November (11)), 954–976. Trebst, C., Jarius, S., Berthele, A., et al., 2014. Update on the diagnosis and treatment of neuromyelitis optica: recommendations of the Neuromyelitis Optica Study Group (NEMOS). J. Neurol. 261 (1), 1–16. Wingerchuk, D.M., Banwell, B., Bennett, J.L., et al., 2015a. International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology 85 (2), 177–189. Wingerchuk, D.M., Banwell, B., Bennett, J.L., Cabre, P., Carroll, W., Chitnis, T., et al., 2015b. International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology 85 (July (2)), 177–189. World Bank Data Team, 2018. New Country Classifications by Income Level: 2018-2019. The Data Blog [Internet][cited 2019 Jan 7]. Available from: https://blogs. worldbank.org/opendata/new-country-classifications-income-level-2018-2019. Zunt, J.R., Kassebaum, N.J., Blake, N., Glennie, L., Wright, C., Nichols, E., et al., 2018. Global, regional, and national burden of meningitis, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 17 (December (12)), 1061–1082.
Declaration of Competing Interest Kathryn Holroyd: none. Andre Vogel: none. Kat Lynch: is an employee of ZoomRx. Brittany Gazdag: is an employee of ZoomRx. Matthew Voghel: is an employee of ZoomRx. Nicholas Alakel: is an employee of ZoomRx. Bryan Patenaude: none. Horacio Chiong-Rivero: none. Farrah Mateen: unrelated to the present work, has consulted for Biogen, Genentech, and Oxford Pharmagenesis, Ltd. Funding This work was supported by The Sumaira Foundation. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.msard.2019.05.013.
50
Contents lists available at ScienceDirect
Multiple Sclerosis and Related Disorders journal homepage: www.elsevier.com/locate/msard
Neuromyelitis optica testing and treatment: Availability and affordability in 60 countries
T
Kathryn Holroyda, Andre Vogela, Kat Lynchb, Brittany Gazdagb, Matthew Voghelb, ⁎ Nicholas Alakelc, Bryan N. Patenauded, Horacio Chiong-Riveroa, Farrah J. Mateena, a
Department of Neurology, Massachusetts General Hospital, 55 Fruit St., Boston, MA 02114, USA ZoomRx, 521 Broadway, New York, NY 10012, USA c ZoomRx, 50 Milk St., Boston, MA 02109, USA d Bloomberg School of Public Health, Johns Hopkins University, 615N. Wolfe Street, Baltimore, MD 21205, USA b
ARTICLE INFO
ABSTRACT
Keywords: Neuromyelitis optica Epidemiology Global health Health economics Immunosuppressive therapy
Background: We characterize the variations in availability and affordability of NMO diagnostic testing and treatment by geographic region and country-level income group. Methods: A structured survey was distributed in English, French, and Spanish in late 2018 to neurologists and other physicians who encounter NMO patients. Results: Respondents (response rate 45%, 64/143 countries contacted) came from all WHO world regions and World Bank country income levels (49% university-based; 13 low-, 16 lower middle-, 16 upper-middle-, and 15 high-income countries). The average cost of an aquaporin-4 antibody (AQP4-Ab) test to a patient globally was 209 USD, and the average cost of NMO treatment per year was 3,819 USD. AQP4-Ab and myelin oligodendrocyte glycoprotein-antibody (MOG-Ab) testing were available in 68% and 38% of all countries. Low-income countries had poor availability of both AQP4-Ab (2/13 countries) and MOG-Ab (1/13) compared to high-income countries (15/15 AQP4-Ab, 13/15 MOG-Ab). Nearly half (48%, 13/27) of African and Eastern Mediterranean countries had access to neither test. Global treatment availability and usage: Azathioprine (88%), rituximab (50%), mycophenolate mofetil (57%), intravenous methylprednisolone (98%), oral prednisone (68%), plasma exchange (78%), intravenous immunoglobulin (72%). Whereas 70–100% of high-income countries’ patients could afford treatment without incurring a catastrophic health expenditure, <10% of low-income country patients could. Most low-income countries (12/13) reported the patient pays for NMO care entirely without public assistance Conclusions: There is a gap in access to diagnostic testing for NMO in non-high-income countries, even in countries where acute and immunosuppressive treatment for NMO are available.
1. Introduction Neuromyelitis Optica (NMO) is found in diverse populations with a prevalence ranging from 0.5 to 4.4 cases per 100,000 people. (CabreraGómez et al., 2009; Asgari et al., 2011; Etemadifar et al., 2015) However, the burden of NMO is likely higher than recognized, especially in resource-limited settings, given variations in knowledge and access to diagnostic tests as well as the misdiagnosis of NMO as multiple sclerosis (MS). (Mealy et al., 2018) NMO differs in treatment from MS, requiring diagnostic distinction at the time of presentation. In 2017, the International Panel for NMO Diagnosis convened to create uniform
diagnostic criteria. (Wingerchuk et al., 2015a) In addition to clinical characteristics, the criteria rely heavily on neurodiagnostic tests, including serum antibody testing and MRI findings. Diagnostic tests for NMO are anecdotally known to be unavailable in several locations globally. In a 2016 study, the aquaporin-4 antibody (AQP4-Ab) test was reported as the most commonly sought after but unavailable test in SubSaharan African postgraduate neurology training programs. (Mateen et al., 2016) NMO is also not a benign condition, with a disease mortality ranging from 9% to 32% in a high-income country, (Mealy et al., 2018) making diagnostic barriers potentially fatal. Early NMO diagnosis and immunosuppressive treatment are
Correspondence to: Department of Neurology, Massachusetts General Hospital, 165 Cambridge Street, #627, Boston, MA 02114, USA. E-mail addresses: [email protected] (K. Holroyd), [email protected] (A. Vogel), [email protected] (K. Lynch), [email protected] (B. Gazdag), [email protected] (M. Voghel), [email protected] (N. Alakel), [email protected] (B.N. Patenaude), [email protected] (H. Chiong-Rivero), [email protected] (F.J. Mateen). ⁎
https://doi.org/10.1016/j.msard.2019.05.013 Received 21 March 2019; Received in revised form 8 April 2019; Accepted 20 May 2019 2211-0348/ © 2019 Elsevier B.V. All rights reserved.
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
essential to reduce both morbidity and mortality for patients. (Pandit et al., 2013; Trebst et al., 2014) In higher income settings, case fatality is thought to have decreased since the introduction of the AQP4-Ab test in 2004. (Collazo et al., 2018) It is unclear if this benefit has reached most populations. We are unaware of global studies that have characterized physicians’ familiarity or the availability and affordability of NMO diagnosis and treatment, although others have examined the characteristics of NMO in lower-income country populations. (Gaye, 2018; Assadeck et al., 2019; Bhigjee et al., 2017) We sought to understand the patterns of diagnosis and treatment of NMO worldwide by surveying a convenience sample of physicians globally who encounter NMO patients.
The survey and introduction were written and distributed in English, French, or Spanish, with the option to request a translation into other languages as the first question of the survey (Appendix A, survey instruments). A gift card for 50USD was offered upon survey completion. 2.5. Data analysis When multiple respondents from a country completed the survey, the first complete submitted response was used in the analysis. Countries were stratified by WHO geographical region (https://www. who.int/about/regions/en/) and World Bank income status from the most recently available year World Bank Data Team (2018). Currencies were converted to USD using the exchange rates on xe.com on November 8, 2018. Affordability analyses were conducted using Stata (v.15.1, College Station, TX).
2. Methods 2.1. Ethics approvals
2.6. Cost analysis
The Partners Healthcare Institutional Review Board exempted this study.
To generate estimates of the affordability of NMO immunosuppressive medications in each country, the distribution of household income was extrapolated. Given limited comparable data on household income by percentile, income distributions were constructed using per capita gross domestic product (measured in purchasing power parity adjusted constant 2011 international$) and the most recent Gini index values, a measure of income inequality, obtained from the World Bank Databank. The precise method for constructing the estimated income distributions has been reported (Harttgen and Vollmer, 2011). In order to adjust the estimated distribution to reflect a household's disposable income, rather than purely individual aggregate income: (1) labor's share of income was assumed to be 2/3 of total income, an estimate stemming from economic theory (Kaldor, 1961; Solow, 1957; Johnson, 1954); (2) it was assumed that there were two wage earners per household; (3) household subsistence was defined as the average number of members per household in each country multiplied by the international annual poverty line. The average number of members per household for each country was obtained from the UN Statistical Database. The international subsistence level was defined as an individual living on $693.50 PPP-adjusted 2011 constant $ per year ($1.90 per day) by World Bank standards (Ravallion et al., 2008). This $693.50 per year is multiplied by the average number of household residents to calculate a measure for household subsistence in constant dollars. We assume that until household income surpasses this bare minimum, there is no disposable income. Household subsistence was subtracted from the previously generated income distribution to calculate household disposable incomes. Affordability was assessed by calculating the cost of one year of NMO treatment as a percentage of annual household disposable income. The WHO defines medical expenditures that are >40% of household disposable income as catastrophic. However, even an expenditure <40% of annual household income is likely catastrophic for most families in the world.
2.2. Survey design and implementation A new survey instrument was designed in 2018 by the authors who include neuroimmunologists, neurology residents, and researchers proficient in survey design for physicians. The survey was piloted on a small group of neuroimmunology-focused neurologists to improve the clarity and sequencing of the questions. The final survey instrument was edited for format, content, and design and had up to 48 multiple choice and fill-in-the-blank questions with contingencies based on prior responses. The goal of the survey was to understand: 1) practice and institution information of the respondent; 2) familiarity with the medical aspects of NMO; 3) imaging available for diagnosing NMO and barriers to its acquisition; 4) diagnostic laboratory tests available for diagnosing NMO and barriers to obtaining these tests; and 5) treatments available for NMO in the acute and chronic phases. The survey was distributed on an independent survey platform, hosted by the company ZoomRx, and available to potential respondents from July to November 2018. 2.3. Respondent selection Countries with >1 million inhabitants and designated as low- or middle-income by the World Bank were prioritized. (Bhigjee et al., 2017) Medical providers thought to be familiar with NMO were identified through several routes: (1) PubMed was searched with MeSH terms, including the country name of interest AND [“neuromyelitis optica,” OR “multiple sclerosis,” OR “neuroimmunology”], and if no findings, the country name AND “neurology;” (2) An online search of neuroimmunologists on faculty at academic hospitals in countries of interest was performed; (3) Delegates of international neuroimmunology committees were located on publicly accessible websites; (4) Presenting authors from countries of interest were found through international neuroimmunology conferences; and (5) Professional contacts of the physician authors’ networks were contacted. In countries where there were no subspecialists of neuroimmunology found, general neurologists were identified. In countries where no neurologists were identified, non-neurologist physicians who would take care of NMO patients were identified.
2.7. Data availability Anonymized survey data will be made available to qualified investigators upon request. 3. Results
2.4. Survey implementation
3.1. Respondents
The response goal was one respondent per country. When a country had a respondent, no further respondents were sought from that country. Respondents were emailed with an introductory message of standardized text requesting their participation in the survey, with a reminder email sent 1 to 4 weeks later if there was no initial response.
Of the 193 UN member states, a physician in 143 was contacted. Responses were received from 60/143 (42%) of countries contacted. Additional respondents from four other countries reported unfamiliarity with NMO and were removed from further analyses (overall response rate 64/143 countries=45%). 45
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
institution. 71% of high-income country respondents practiced at an academic institution, while 39% of low-income and 40% of lowermiddle-income country respondents did so. By WHO region, 67% of European respondents reported practicing at an academic institution compared to only 39% of respondents from African countries.
Table 1 Respondent characteristics. n = 60 Degree of experience with NMO diagnosis and treatment: Familiar, but never diagnosed Familiar, have diagnosed, but never treated Fairly familiar, have diagnosed, treated, but not often Very familiar, diagnose and treat regularly Setting of primary practice location: Rural (<50,000 residents) Smaller city (100,000 to 1 million residents) Larger city (100,000 to 1 million residents) Urban area (>1 million residents) World Bank income status of country of respondent: Low income (L) Lower-middle income (LM) Upper- middle income (UM) High income (H) WHO region of respondent: Africa (Af) Americas (Am) Eastern Mediterranean Region (EM) Europe (Eu) South-East Asia (SEA) Western Pacific (WP) Primary treated age group: Children (0–17 years) Adults (18+ years) Both children and adults Confidence treating NMO. 1 = Not at all, 7 = extremely confident 1–3 4 5–7
7 (12) 4 (7) 23 (38) 26 (43)
3.2. Diagnostic tests Across all country-level respondents, AQP4-Ab and MOG-Ab were present in 41 (68%) and 23 (38%) countries respectively. 15 of 15 (100%) high-income countries versus 2 of 13 (15%) low-income countries reported AQP4-Ab availability. 13 of 15 (87%) high-income countries versus 1 of 13 (8%) low-income countries reported MOG-Ab availability. 13 of 27 (48%) respondents in Africa and the Eastern Mediterranean regions reported that the AQP4-antibody test is unavailable. In contrast, only 6 of 33 (18%) respondents from countries of any income level in the pooled Americas, Europe, South-East Asia, and the Western Pacific reported that the AQP4-Ab test is unavailable. Of the 41 respondents who reported availability of the AQP4-Ab test, 20 (49%) stated they had access to the test without issue. Similarly, of the 23 respondents who reported availability of the MOG-Ab test, 13 (57%) reported having access to the test without issue. However, when respondents were aggregated by WHO world region, only 4 of 14 (29%) respondents from Africa and the Eastern Mediterranean reported access to the AQP4-Ab test and 2 of 8 (25%) reported access to MOG-Ab test without issue (Table 2). Most respondents had access to the other diagnostic tests that support the diagnosis of NMO. Fifty respondents (83%) reported availability of a CT scanner on site, 42 (68%) had an MRI scanner on site, 50 (83%) could perform a lumbar puncture on site, 36 (60%) could access visually evoked potentials on site, and 34 (57%) could access optical coherence tomography on site. Only CT scans and lumbar punctures were available on site in most low-income countries (Fig. 2). Countries in Africa and the Eastern Mediterranean regions reported that MRI, CT, and lumbar puncture were available at roughly similar rates as countries in the Americas, Europe, South-East Asia, and the Western Pacific; however, they reported significantly lower rates of availability of visual
1 (2) 2 (3) 14 (23) 43 (72) 13 16 16 15
(22) (27) (27) (25)
15 (25) 10 (17) 12 (20) 10 (17) 6 (10) 7 (12) 2 (3) 50 (83) 8 (13) 4 (7) 10 (17) 46 (77)
Of the 60 respondents who completed the survey, >90% of respondents had at least an average level of confidence in their care of people with NMO (Table 1). Respondents from all world regions and country-income level strata completed the survey (Fig. 1). Most respondents practiced in an urban area of >1 million inhabitants with only one respondent practicing in a rural area of <50,000 inhabitants. Approximately half (49%) of respondents practiced at an academic
Fig. 1. Map of responding countries by World Bank income status. 46
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
Table 2 Availability, access, and barriers to NMO diagnosis and treatment across income levels and world regions. Availability and access of diagnostic tests
H (n = 15)
UM (n = 16)
LM (n = 16)
Availability of AQP4-Ab diagnostic 15 (100) 13 (81) 11 (69) test Availability of MOG-Ab test 13 (87) 4 (25) 5 (31) No availability of either AQP4- or 0 (0) 3 (19) 5 (31) MOG-Ab tests Where AQP4-Ab test available, 11/15 (73) 4/13 (29) 3/11 (23) access without issue (n = 44) Where MOG-Ab test available, 10/13 (77) 1/4 (25) 1/5 (17) access without issue (n = 24) On-site and referral availability of common NMO diagnostic tools CT scan 13 (87) 16 (100) 15 (94) Lumbar puncture 14 (93) 16 (100) 13 (81) Optical Coherence Tomography 14 (93) 13 (81) 7 (56) MRI 15 (100) 16 (100) 14 (88) Visually Evoked Potentials 14 (93) 12 (75) 11 (69) Availability and use of long-term NMO treatments Azathioprine 13 (87) 15 (94) 16 (100) Mycophenolate mofetil 13 (87) 9 (56) 9 (56) Rituximab 12 (80) 9 (56) 6 (38) Challenges to treatment of NMO Inability to consistently obtain 1 (7) 8 (50) 8 (50) medicine supplies Cost of medication to health sector 6 (40) 11 (69) 8 (50) Cost of medication to patient 1 (7) 11 (69) 12 (75) Practice lacks access to preferred 4 (27) 6 (38) 6 (38) medications
L (n = 13)
Af (n = 15)
Am (n = 10)
EM (n = 12)
Eu (n = 10)
SEA (n = 6)
WP (n = 7)
2 (15)
7 (47)
7 (70)
7 (58)
9 (90)
5 (83)
6 (86)
1 (8) 11 (85)
3 (20) 8 (53)
3 (30) 3 (30)
5 (42) 5 (42)
7 (70) 1 (10)
1 (17) 1 (17)
4 (57) 1 (14)
2/2 (100)
2/7 (29)
3/7 (43)
2/7 (29)
7/9 (78)
2/5 (40)
4/6 (67)
1/1 (100)
1/3 (33)
1/3 (33)
1/5 (20)
6/7 (86)
1/1 (100)
3/4 (75)
8 9 5 7 5
(62) (69) (38) (54) (38)
11 (73) 12 (80) 6 (40) 10 (67) 5 (33)
10 (100) 10 (100) 9 (90) 10 (100) 8 (80)
10 (83) 10 (83) 6 (50) 10 (83) 10 (83)
9 (90) 10 (100) 9 (90) 10 (100) 9 (90)
6 5 5 5 4
6 5 6 7 6
9 (69) 3 (23) 3 (23)
11 (73) 7 (47) 2 (13)
9 (90) 6 (60) 6 (60)
12 (100) 5 (42) 7 (58)
9 (90) 9 (90) 8 (80)
6 (100) 3 (50) 3 (50)
6 (86) 4 (57) 4 (57)
9 (69)
11 (73)
4 (40)
5 (42)
1 (10)
3 (50)
2 (29)
5 (38) 11 (85) 9 (69)
7 (47) 13 (87) 11 (73)
7 (70) 5 (50) 4 (40)
8 (67) 8 (67) 7 (58)
4 (40) 1 (10) 1 (10)
2 (33) 5 (83) 1 (17)
2 (29) 3 (43) 1 (14)
(100) (83) (83) (83) (67)
(86) (71) (86) (100) (86)
Note: H = High-income countries, UM = Upper middle-income countries, LM = Lower middle-income countries, L = Low-income countries. Af = Africa, Am = Americas, EM = Eastern-Mediterranean, Eu = Europe, SEA = South-East Asia, WP = Western Pacific.
Fig. 2. Availability and accessibility of non-antibody diagnostic tests by A) World Bank income status and B) WHO World Region of responding countries. 47
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
Fig. 3. Availability and accessibility of long-term NMO treatments by A) World Bank income status and B) WHO World Region of responding countries.
evoked potentials and optical coherence tomography. Respondents from 28 countries reported an estimate of the price of AQP4-Ab testing. The average reported price for the AQP4-Ab test was 209.29USD (median 91 USD).
3.4. Perceived barriers to NMO care The most commonly reported challenge was the cost of immunosuppressive medication to the patient, followed by an inability to obtain a supply of medications, and the cost of the medication to the health sector. A minority (7, 12%) of respondents reported never having experienced challenges treating NMO. With the exception of respondents from high-income countries, the most commonly reported challenge in treating NMO was the cost of medicines to the patient. A majority (11, 85%) of low-income countries found patient-borne costs were a challenge in treatment, while only 1 of 15 (7%) high-income countries reported the same. For respondents from high-income countries, the greatest challenges in treating NMO were the cost to the health sector, inadequate patient understanding of the disease, and patient non-adherence. 9 (69%) respondents from lowincome countries reported lack of access to medications as a barrier to treatment. In contrast, only 4 (27%) respondents from high-income countries reported that care is impeded by a lack of access to medications.
3.3. Treatment Access to treatment for acute NMO relapses was high: 59 (98%) respondents reported the availability of intravenous methylprednisolone, 41 (68%) of oral steroids, 47 (78%) of plasma exchange, and 43 (72%) of intravenous immunoglobulin. Availability of acute NMO treatments increased with a country's World Bank income status from low to high. Access to long term NMO immunosuppressive treatments was variable. Most (n = 53 of 60, 88%) respondents stated that azathioprine is both available and used. Use of mycophenolate mofetil and rituximab increased with increasing country-level income group. Mycophenolate mofetil and rituximab were available in 3 low-income countries (23%) compared to high-income countries where 13 (87%) and 12 (80%) had availability of mycophenolate mofetil and rituximab respectively (Fig. 3). For patients in whom it is impossible to distinguish between NMO and MS at initial presentation based on available in-country resources, there was variation in treatment approach. Clinicians would treat as NMO (25, 42%), MS (16, 27%), or not treat for either disease (19, 32%). Of those who would treat a patient as MS when uncertain on the diagnosis, half of the respondents (8, 50%) stated they would consider azathioprine for initial treatment. Other frequent medication choices included interferon beta-1a, glatiramer acetate, natalizumab, and rituximab.
3.5. Affordability The average annual cost to a typical patient for NMO treatment was estimated at 3818.56 USD (n = 52 responding countries). 21 (35%) respondents stated that a patient's costs are not typically covered by public health insurance, 18 (30%) stated that a patient's costs are partially covered, and 20 (33%) stated that a patient's costs are covered in full (n = 1 no answer). All 15 high-income countries reported that patients’ treatment costs were covered partially or fully by public insurance. In contrast, 12 (92%) low-income countries reported that a patient's costs are not covered at all by public insurance, with one low48
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
Fig. 4. Predicted populations percentages able to afford 1 year of NMO treatment by expending <40% or <10% of annual household income.
consensus criteria for NMO rely heavily on the availability of MRI and blood tests which may be difficult to obtain and require the mailing out of blood samples to laboratories out of country with prolonged waiting time for results. For all countries to reach the same standard of care for the diagnosis of NMO, more universal access to diagnostic tests is necessary. With a more affordable, accessible point of care diagnostic test for NMO, for instance, the ability to diagnose and therefore treat NMO would likely be greatly increased internationally. Since the NMO diagnostic criteria are understandably dependent on the sensitive and specific biomarkers that are available, it is debatable whether the criteria should be modified for level of certainty in low-resource settings versus the diagnostic tests themselves should be made more available, accessible, and affordable. Presently, many patients’ out of pocket costs for diagnosis can far exceed the cost of treatment with azathioprine. For health systems that do not have access to or cannot afford diagnostic testing, the question of treating the disease as NMO or as multiple sclerosis remains unsettled. Although we hypothesized that access to both diagnosis and treatment are limited in more resource-limited settings, our respondents most often reported limitations in NMO diagnosis. To underscore this dilemma, we found that 69% of low-income country respondents had access to azathioprine while only 15% reported access to AQP4-Ab and 8% access to MOG-Ab testing. The ability to treat patients using prednisone and azathioprine was relatively higher than expected. Since azathioprine is available on the WHO's Essential Medicines Lists for other indications, the use of azathioprine is usually possible. Meanwhile, the treatment landscape for NMO continues to expand in high-income countries. For instance, one novel therapy, eculizumab, is currently estimated to cost ∼500,000 USD per patient per year (Pittock et al., 2013; Dinerstein, 2017). For this same price, we estimated that >2000 AQP4-Ab tests could be performed, using the average reported cost in our study of 209.29 USD. There are several limitations to our study. The response rate is 45%. Although this is low compared to the number of countries we hoped to reach, it is similar to other published surveys reaching lower income countries (McLane et al., 2015) and includes a fairly even representation across geographic and country income levels. Countries in lower
income country reporting that a patient's costs are partially covered. In low-income countries for which disposable household income assessment was possible due to available demographic data, none of the patient population could afford the cost of 1 year of NMO medication by spending <10% of disposable income. Less than 10% of the population in any low-income country (often 0), could afford 1 year of NMO medication without incurring catastrophic (>40% of household disposable income) health expenditures. By contrast, 70 to 100% of the population in high-income countries analyzed could afford one year of NMO immunosuppression without incurring catastrophic health expenditures (Fig. 4). 4. Discussion We present data from a global convenience sample that synthesizes practitioners’ ability to diagnose and treat NMO across geographic and economic zones. Whereas prior reports of a global nature have considered the landscape of neurological care by the type of diagnostic tests available (McLane et al., 2015) or the global burden of a particular neurological disease, (Zunt et al., 2016; Stovner et al., 2018; Feigin et al., 2017) few studies have combined these two approaches. Our study allows an improved understanding of the delivery of care for a neuroimmunological disease and how the availability of testing and treatment varies across broad world regions. Since NMO has relatively recently been defined by its distinct biomarkers (Wingerchuk et al., 2015b) our study also sheds light on the pace of uptake of new technologies and treatments for a neurological disease around the world when they originate from a high-income setting. Our study both emphasizes existing trends in health systems and sheds new light on the burden neurologists face when encountering a less common neurological disorder. Our results, although mainly descriptive, could lead to several important interventions that improve the lives of patients with NMO and enable the practitioners who encounter them. First, we identified several countries where neurology practitioners, whether neurologists or not, were unfamiliar with NMO and could be invited to ongoing educational initiatives to increase NMO awareness. Second, diagnostic 49
Multiple Sclerosis and Related Disorders 33 (2019) 44–50
K. Holroyd, et al.
income settings may have been more motivated to reply, because they have recognizable needs and fewer available respondents. It is possible that we overlooked eligible clinicians in some countries. Because we reached out to personal contacts, our respondents’ demographic characteristics may be biased towards those in a similar practice type as the authors, i.e. in academic settings. Our results are subject to selection bias. Since our a priori response goal was to achieve one key informant in a country, we did not aim to have multiple respondents which could have provided a more complex picture of NMO care within each country. Although survey translation was offered, the survey was initially distributed in English, with French and Spanish translations made available following requests. This limited our response rate to people who spoke at least basic levels of one of these languages. The use of a web-based survey may have led to selection bias towards respondents with better access to other resources; however, we could not find any other way to reach so many practitioners globally in such a short period of time. Responses may have been inaccurate and questions may have been misunderstood. We report trends and patterns, but do not identify causal associations. It is well-recognized that countries have significant regional variability and sub-national inequities. Our respondents, being disproportionately in urban, academic centers, may provide better than usual care in a country. Some countries have insufficient baseline household information to calculate affordability because they are too poor or fragile. Given this, the affordability of NMO diagnosis and treatment may have been over-estimated. Much work remains to be done to increase the availability and affordability of diagnostic tests for neuroimmunological diseases like NMO in lower income countries. Although treatments for acute disease relapses are generally available, very limited access to testing can be used to direct management. Increasing international education programs and collaborating with established non-governmental organizations and pharmaceutical companies could help increase awareness for NMO and access to existing diagnostic tests, or even contribute to development of a point-of-care test.
References Asgari, N., Lillevang, S.T., Skejoe, H.P., Falah, M., Stenager, E., Kyvik, K.O., 2011. A population-based study of neuromyelitis optica in Caucasians. Neurology 76 (18), 1589–1595. Assadeck, H., Toudou Daouda, M., Adehossi Omar, E., Mamadou, Z., Hassane Djibo, F., Douma Maiga, D., 2019. Inflammatory demyelinatiung diseases of the central nervous system in Niger. Rev. Neurol. (Paris). 175 (4), 261–268. Bhigjee, A.I., Moodley, A.A., Roos, I., Wells, C.L., Ramdial, P., Esser, M., 2017. The neuromyelitis optica presentation and the aquaporin-4 antibody in HIV-seropositive and seronegative patients in KwaZulu-Natal, South Africa. South Afr. J. HIV Med. 18 (1), 684. Cabrera-Gómez, J.A., Kurtzke, J.F., González-Quevedo, A., Lara-Rodríguez, R., 2009. An epidemiological study of neuromyelitis optica in Cuba. J. Neurol. 256 (1), 35–44. Collazo, I.M., Weinshenker, B., Nasr, Z., Lopez, A., 2018. Long term mortality in neuromyelitis optica spectrum disorder (P6.414). Neurology 90 (April(15 Supplement)). Dinerstein, C., 2017. Why is Soliris the Most Expensive Drug in the US? American Council on Science and Health [Internet][Accessed March 7, 2019]. Available from: https:// www.acsh.org/news/2017/05/27/why-soliris-most-expensive-drug-us-11333. Etemadifar, M., Nasr, Z., Khalili, B., Taherioun, M., Vosoughi, R., 2015. Epidemiology of neuromyelitis optica in the world: a systematic review and meta-analysis. Mult. Scler. Int. 2015, 174720. Feigin, V.L., Abajobir, A.A., Abate, K.H., Abd-Allah, F., Abdulle, A.M., Abera, S.F., et al., 2017. Global, regional, and national burden of neurological disorders during 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet Neurol. 16 (November (11)), 877–897. Gaye, N.M., Fall, M., Diop, A.M., 2018. Neuromyelitis optica spectrum disorders (NMOSD) in a Sub-Saharan Africa country: a preliminary study of sixteen Senegalese cases. Mult. Scler. Relat. Disord. 27, 179–183. Harttgen, K., Vollmer, S., 2011 November. United Nations Development Programme. Human Development Reports. Research Paper 2011/13. Johnson, D.G., 1954. The functional distribution of income in the United States, 18501952. Rev. Econ. Stat. 36 (2), 175–182. Kaldor, N., 1961. Capital accumulation and economic growth. The Theory of Capital. Macmillan & Co Ltd. Mateen, F.J., Clark, S.J., Borzello, M., Kabore, J., Seidi, O., 2016. Neurology training in sub-Saharan Africa: a survey of people in training from 19 countries. Ann. Neurol. 79 (6), 871–881. McLane, H.C., Berkowitz, A.L., Patenaude, B.N., et al., 2015. Availability, accessibility, and affordability of neurodiagnostic tests in 37 countries. Neurology 85 (18), 1614–1622. Mealy, M.A., Kessler, R.A., Rimler, Z., et al., 2018. Mortality in neuromyelitis optica is strongly associated with African ancestry. Neurol. Neuroimmunol. Neuroinflamm. 5 (4), e468. Pandit, L., Mustafa, S., Kunder, R., Shetty, R., Misri, Z., Pai, S., 2013. Optimizing the management of neuromyelitis optica and spectrum disorders in resource poor settings: experience from the Mangalore demyelinating disease registry. Ann. Indian Acad. Neurol. 16 (4), 572–576. Pittock, S.J., Lennon, V.A., McKeon, A., et al., 2013. Eculizumab in AQP4-IgG-positive relapsing neuromyelitis optica spectrum disorders: an open-label pilot study. Lancet Neurol. 12 (6), 554–562. Ravallion, M., Chen, S., Sangraula, P., 2008. Dollar a Day Revisited (Report). The World Bank, Washington DC. Solow, R.M., 1957. Technical change and the aggregate production function. Rev. Econ. Stat. 39 (August (3)), 312. Stovner, L.J., Nichols, E., Steiner, T.J., Abd-Allah, F., Abdelalim, A., Al-Raddadi, R.M., et al., 2018. Global, regional, and national burden of migraine and tension-type headache, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 17 (November (11)), 954–976. Trebst, C., Jarius, S., Berthele, A., et al., 2014. Update on the diagnosis and treatment of neuromyelitis optica: recommendations of the Neuromyelitis Optica Study Group (NEMOS). J. Neurol. 261 (1), 1–16. Wingerchuk, D.M., Banwell, B., Bennett, J.L., et al., 2015a. International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology 85 (2), 177–189. Wingerchuk, D.M., Banwell, B., Bennett, J.L., Cabre, P., Carroll, W., Chitnis, T., et al., 2015b. International consensus diagnostic criteria for neuromyelitis optica spectrum disorders. Neurology 85 (July (2)), 177–189. World Bank Data Team, 2018. New Country Classifications by Income Level: 2018-2019. The Data Blog [Internet][cited 2019 Jan 7]. Available from: https://blogs. worldbank.org/opendata/new-country-classifications-income-level-2018-2019. Zunt, J.R., Kassebaum, N.J., Blake, N., Glennie, L., Wright, C., Nichols, E., et al., 2018. Global, regional, and national burden of meningitis, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 17 (December (12)), 1061–1082.
Declaration of Competing Interest Kathryn Holroyd: none. Andre Vogel: none. Kat Lynch: is an employee of ZoomRx. Brittany Gazdag: is an employee of ZoomRx. Matthew Voghel: is an employee of ZoomRx. Nicholas Alakel: is an employee of ZoomRx. Bryan Patenaude: none. Horacio Chiong-Rivero: none. Farrah Mateen: unrelated to the present work, has consulted for Biogen, Genentech, and Oxford Pharmagenesis, Ltd. Funding This work was supported by The Sumaira Foundation. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.msard.2019.05.013.
50